Your new Google Ads campaign spent $572 since Monday. It's Wednesday afternoon. You have exactly one lead, and it was your own test conversion from the setup process. The client’s email lands in your inbox: “Checking in on the new campaign… Is this working? The cost per lead seems high.”

The gut reaction is to panic. The second reaction is to throw out a comforting cliché like, “We need to let the algorithm learn!” or, “Let’s give it 7 more days.”

Both reactions are wrong.

The decision to kill a campaign or let it run has nothing to do with a specific number of days. It has everything to do with diagnosing the type of problem you're facing. Some problems are fatal. Others are just expensive growing pains.

The 7-Day Rule Isn't a Rule, It's a Starting Point

Platforms like Google and Meta have a “learning phase.” This is a real, defined period where the ad delivery system is actively exploring who to show your ads to. It’s gathering data on clicks, user characteristics, and conversions to build a profile of your ideal customer. During this phase, performance is volatile. Your Cost-Per-Lead (CPL) might swing from $50 to $500 and back in a matter of days.

Killing a campaign mid-learning-phase because you don’t like the numbers is like pulling a cake out of the oven after 10 minutes because it’s not fluffy yet. You reset the entire process.

However, this doesn't mean you just blindly trust the process for a set number of days. Waiting 7 days is pointless if the campaign's foundation is cracked. The key is to know what to look for. Are you seeing signs of a slow start, or signs of a fundamental, unfixable flaw? That’s the entire game.

Red Flags vs. Amber Flags: Learning to Tell the Difference

You don’t need more time. You need a better diagnostic framework. We split early campaign problems into two categories: Red Flags (kill it now) and Amber Flags (wait, but watch closely).

Red Flags: Shut It Down Immediately

A red flag is a problem with the campaign’s core structure or setup. Letting it run longer won't fix it; it will only waste more money.

• Totally Broken Tracking: You launch, spend $500, and then realize your GTM container wasn't published or the Meta pixel event is firing on the wrong page. You are flying completely blind. There is no conversion data to optimize for. This is a non-negotiable kill switch. Pause everything, fix the tracking, test it yourself, then relaunch. Don’t spend another dollar without a working speedometer.
• Zero Impressions/Delivery: 24 hours have passed and you've spent $0.00. Don't just wait. Something is wrong. Is the ad disapproved? Is there a billing failure? Are your bids absurdly low or your audiences impossibly small? This is a technical problem, not a performance problem. Find the roadblock and fix it.
• Catastrophic Misfire on Targeting: You sell high-end HVAC systems. Your search terms report after one day is full of queries like “free AC repair manual” or “how to fix fan myself.” Or, you sell B2B software and you realize you left the Google Display Network active and spent $1,000 on mobile game apps before 10 AM. This isn’t a learning issue. This is a fundamental targeting failure. Kill the campaign, refine your keywords (hello, negative keywords), and check your network settings. Rebuild it correctly.

Amber Flags: Wait & Diagnose

An amber flag is a performance issue, not a structural one. The campaign is working, just not as well as you’d like. This is where patience, paired with active diagnosis, pays off.

• High Cost-Per-Lead (CPL): This is the most common reason for early panic. The CPL is 2x your target. It’s painful, but you have leads. The system is finding people who convert. Now your job is to figure out if it’s an expensive trend or a temporary average. Is the CPL dropping day over day? Are the leads qualified? High CPL is a problem to be solved with optimization, not a reason to scrap a campaign that’s generating data.
• Low Click-Through-Rate (CTR): Your impressions are high, but nobody's clicking. This means your targeting might be okay, but your ad creative or offer isn't compelling. Does this justify killing the whole campaign? Absolutely not. It justifies duplicating the ad set and testing new copy or a new image. The underlying structure could be sound.
• High CTR, Low Conversion Rate: This is one of my favorite “problems” to have. Lots of people are clicking your ad, but nobody is converting on the landing page. This is a gift. It means the ad is doing its job perfectly—it’s attracting the right people and compelling them to act. The problem isn't the campaign; it's the post-click experience. Don't you dare touch that campaign. Go fix the landing page: simplify the form, clarify the headline, improve the page speed.

A Tale of Two Campaigns

Theory is fine. Let's make this concrete with two real-world scenarios.

Scenario 1: Apex Roofing (Amber Flag)

A local roofing company launches a campaign for emergency roof repairs. Their target CPL is $250. After five days, they’ve spent $1,200 and generated three leads, for an average CPL of $400.

The client sees the $400 CPL and wants to pull the plug. But we dive in. The search query report shows exactly what we want: “emergency roofer near me,” “leaking roof repair houston,” “storm damage roofer.” The targeting is perfect. The three leads are all from homeowners with legitimate, urgent problems. The issue is simply cost. This is a classic Amber Flag. The campaign is working, it's just expensive right now. We advise staying the course for another 5-7 days to let the algorithm optimize and collect more conversion data. The tradeoff is spending more money now for a potentially more efficient campaign later. We also start looking at bid adjustments and testing new ad copy to improve CTR, but we don't kill the core campaign.

Scenario 2: FinTech Advisors (Red Flag)

A wealth management firm launches a campaign targeting high-net-worth individuals. Their target CPL is $350. After five days, they’ve spent $2,000 with zero qualified leads. They got two form fills, but one was from a student researching a paper and the other was an obvious spam bot.

We look at the search query report. Because they used the broad match keyword investing, they are showing up for queries like “best reddit investing advice,” “robinhood alternative,” and “how to start investing with $100.” The traffic they are buying has no commercial intent and is completely misaligned with their service. This is a textbook Red Flag. Letting this run another day, let alone seven, is just setting money on fire. The campaign's foundation—its keyword strategy—is fundamentally broken. We kill it immediately, take the loss, and rebuild from scratch with tight phrase/exact match keywords and an extensive negative keyword list.

Your Pre-Launch Checklist to Avoid Early Panic

Most of this frantic early-stage decision-making can be avoided by doing the work up-front.

1. Verify Tracking End-to-End: Before you activate the campaign, go through the user flow yourself. Click an ad preview, fill out the form, and make sure the conversion shows up in the ad platform and your CRM. Do it twice.
2. Set a Tight Initial Budget: Don't launch a brand-new, unproven campaign on a $1,000/day budget. Start with $100 or $200. Limiting the initial daily spend gives you time to gather data without risking the entire month's budget in 48 hours.
3. Know Your Breakeven Numbers: Before you spend a dollar, you should know your target CPL/CPA. Not a vague goal, but the actual number. What can you afford to pay for a lead and still be profitable? This number is your North Star. Without it, you're just reacting to sticker shock.
4. Have a Hypothesis: Define what success looks like beyond just cost. State it clearly: "My hypothesis is that by targeting these specific keywords, we can generate leads under $150 with a conversion rate of 5%." Now you're not just waiting, you're proving or disproving a specific thesis.

The knee-jerk reaction to high initial costs is emotional, not analytical. The difference between a pro and an amateur is the ability to ignore the emotion, categorize the problem, and act accordingly. Is this a tune-up or a cracked engine block? One requires a wrench and some patience. The other requires you to cut your losses and walk back to the workshop. The "7-day rule" doesn't tell you which is which; only a rigorous diagnosis of the data can.